|
|
||||||||
1


* Goldyne Savad Institute of Gene Therapy,
Liver Unit, and
Department of Pathology Hadassah University Hospital, Jerusalem, Israel; and
§ I. Medizinische Klinik, Abteilung Pathophysiologie, Johannes Gutenberg Universitat, Mainz, Germany
1Correspondence: Goldyne Savad Institute of Gene Therapy; Hadassah University Hospital; Jerusalem 91120; Israel. E-mail: galun{at}md2.huji.ac.il
| ABSTRACT |
|---|
|
|
|---|
, and
TGF-
). Many IL-6 effects depend on a naturally occurring soluble
IL-6 receptor (sIL-6R). Here, the chimeric protein hyper-IL-6,
constructed from the human IL-6 protein fused to a truncated form of
its receptor, was found to have superagonistic IL-6 properties, and as
such, enhanced liver cell regeneration. Hyper-IL-6 reversed the state
of hepatotoxicity and enhanced the survival rates of rats suffering
from fulminant hepatic failure after D-galactosamine administration.
The hyper-IL-6 protein has a significant potential for use in the
treatment of severe human liver diseases.Galun, E., Zeira, E., Pappo,
O., Peters, M., Rose-John, S. Liver regeneration induced by
a designer human IL-6/sIL-6R fusion protein reverses severe
hepatocellular injury.
Key Words: interleukin 6 chimeric protein hyper-IL-6 cytokines hepatotoxicity liver failure
| INTRODUCTION |
|---|
|
|
|---|
60% (1)
One of the causes of death by FHF is liver cell apoptosis or hepatocyte
death through necrosis, as seen in the case of acute hepatitis A virus
infection (6)
. Thus, FHF might be cured by factors that
could induce liver cell regeneration (7
8
9
10)
. It has
recently been reported that liver cell proliferation is also enhanced
by interleukin 6 (IL-6) (11
, 12)
. IL-6 binds to cells
through the IL-6 receptor
(IL-6R
, gp80), thus facilitating its
interaction with a second IL-6 receptor molecule, IL-6Rß (gp130).
IL-6R
can also be found as a soluble protein that is thought to be
generated either by limited proteolysis of the membrane-bound receptor
(13)
or by translation of an alternatively spliced mRNA
(14)
. In a process called trans-signaling, soluble
IL-6R (sIL-6R) has been shown to sensitize target cells
(15)
and to cause cells that do not express membrane-bound
IL-6R
to be responsive to IL-6 (16
, 17)
. Marked
hepatocellular hyperplasia is seen in IL-6/sIL-6 receptor (IL-6/sIL-6R)
double transgenic mice, but not in single transgenic IL-6 mice,
suggesting that sIL-6R recruits IL-6-unresponsive hepatocytes to
proliferation (18
, 19)
. Recently, a fusion protein called
hyper-IL-6 was constructed, consisting of human IL-6 and the human
sIL-6R connected by a flexible peptide chain (20)
.
Hyper-IL-6 was shown to exhibit a high activity level on
gp130-expressing cells both in vitro (20)
and
in vivo (21)
. It appears that the hyper-IL-6
fusion protein acts as a superagonist by simulating the interaction
between IL-6 and sIL-6R
(20)
.
In the present study, we have assessed the effects of hyper-IL-6 in inducing hepatocellular proliferation and liver cell regeneration in vivo, with the aim of determining its potential therapeutic value.
| MATERIALS AND METHODS |
|---|
|
|
|---|
D-gal rat nonlethal severe hepatotoxicity model
We slightly modified (see below) the male Fischer rat model in
which a state of nonlethal severe hepatotoxicity was created by the
application of D-galactosamine (D-gal). To determine the level of
hepatotoxicity, we used the following clinical biochemical criteria:
1) blood glucose levels, since hypoglycemia is induced by
the failure of liver homeostatic mechanisms; 2) bilirubin
levels, since jaundice often accompanies acute liver injury;
3) the blood level of alanine transferase (ALT), since such
hepatocyte cytosolic enzyme transaminases are released into the blood
when liver cells are injured. In addition, since most blood coagulation
factors are produced in the liver, in severe acute liver disease the
production of coagulation factors is typically reduced. Coagulation
factors V and VII have a short t1/2 of less than
8 h, so we chose to measure their serum levels to follow the
course of the liver injury. Before we administered D-gal, the results
of the liver function tests were as follows (mean ±
SD, n=8): glucose 84±13 mg%;
coagulation factor V 119± 41%; coagulation factor VII 93±10%; ALT
55±4 I.U., and bilirubin 4± 0 µM/l.
For these studies we used male Fisher rats weighing 150200 g (Harlan,
Animal Breeding Center, Jerusalem, Israel) that were fed a standard
diet of rat chow and tap water and housed in standard facilities at
room temperature of 25°C with a 12 h day/night cycle. After the
rats had fasted for 12 h, acute liver damage was induced by the
intraperitoneal (i.p.) administration of 100500 mg/kg body weight
doses of D-gal (Sigma Chemical Co., G0264, Israel) dissolved in 0.9%
NaCl and adjusted with 1N NaOH to pH 6.8. After the injection of D-gal,
the rats were fasted for an additional 12 h, but were provided
with water containing 10% glucose ad libitum to maintain
the blood glucose level. To determine the optimal dose of D-gal for
inducing severe but nonlethal hepatotoxicity, five groups of two rats
each received a single i.p. dose of 100 mg, 200 mg, 300 mg, 400 mg, or
500 mg/kg body weight injection of D-gal. The results of a D-gal dose
response experiment (Fig. 1
) revealed that these rats developed the hepatotoxic effect after the
i.p. administration of 300 mg/kg of D-gal. We used this dosage in later
experiments designed to assess the effects of hyper-IL-6, IL-6, and
glucose treatments (Fig. 2
); see below.
|
|
D-gal male Fischer rat FHF model
The dosage of D-gal required for 100% mortality of male rats
was revealed in a dose response experiment in which male Fischer rats
were injected i.p. with D-gal at dosages of 1, 1.2, or 1.4 gm/kg body
weight (Fig. 3
). In similar experiments on female Fischer rats (animal weight 150170
g), the dose response survival experiment was conducted with doses
range between 100 and 1400 mg/kg. All the female animals died when
treated with D-gal at a dose of 300 mg/kg or higher.
|
Assessment of liver cell proliferation by 5-bromo-2'-deoxyuridine
(BrdU) incorporation
We assessed liver regeneration as reflected by cell
proliferation measured by BrdU immunohistological staining (see Fig. 5
). One and 2 h before killing animals, an i.p. injection of a PBS
solution of the thymidine analog 5-bromo-2'-deoxyuridine (Sigma
Chemical Co.) was administered at a dose of 50 mg/kg body weight.
Livers were harvested and fixed in 4% formaldehyde buffer. An
automated tissue processor was used for fixation, followed by liver
embedding in paraffin. Tissue sections (45 µ) were cut on a
microtome and adhered to poly-L-lysine-coated glass slides. Staining of
fixed tissue samples was carried out using an antibody to BrdU (Zymed,
San-Francisco, Calif.), enabling the proliferating cells (red nuclei)
to be distinguished from others (blue nuclei). The immunohistochemical
study was performed according to the manufacturers instructions
(Zymed BrdU labeling and detection kit).
|
Thioacetamide (TAA) FHF rat model
A TAA dose response study was conducted with male
Sprague-Dawley rats (weight 180220 g) that received TAA in 5 ml
saline administered i.p. daily for 3 consecutive days. To prevent
hypoglycemia and hypokalemia, at 12 h intervals after the initial
TAA injection the animals were treated with subcutaneous injections of
25 ml/kg body weight of 5% dextrose/0.45% saline containing 20 mEq/l
KCl; i.p. administration of TAA at 300 mg/kg body weight induced death
from FHF in 100% of these male Sprague-Dawley rats.
Assessing liver cell apoptosis with the TUNEL assay
Male Fischer rats were treated with D-gal at a dosage of 300
mg/kg body weight; 7 h later they were treated i.p. with either
IL-6 at 80 µg/rat or HIL-6 at 8 µg/rat. Two days after the D-gal
injections, the livers were harvested fixed in 4% formaldehyde and
paraffin embedded. Apoptosis was examined in 4 µm sections, as
assayed by direct immunoperoxidase to detect digoxigenin-labeled
genomic DNA. As a positive control we used sections treated with DNase
I to nick all DNA (1 µg/ml, determined after a preliminary dilution
experiment); for a negative control, we used sections that were only
immersed in a Tdt buffer containing 3 mM of biotin dUTP. The TUNEL
staining was prepared according to modified protocol of Boehringer
Mannheim, (Indianapolis, Ind.). Visual images of TUNEL-stained slides
were captured with a digital camera attached to a microscope. Apoptotic
cells were identified at the color threshold set (dark brown) for
3',3-diaminobenzidine tetrahydrochloride.
| RESULTS |
|---|
|
|
|---|
Hyper-IL-6 relieves nonlethal severe hepatotoxicity in a rat model
Using male rats treated with D-gal to induce severe
hepatotoxicity, we compared the effects of hyper-IL-6 to those of human
IL-6 or of glucose. Seven hours after the i.p. injection of D-gal, we
administered hyper-IL-6, IL-6, or glucose (see legend to Fig. 2
). The
failure of liver homeostatic mechanism is known to be accompanied by
various physiological changes, which we monitored in D-gal-treated
rats. Jaundice is a common manifestation of acute liver injury, so we
measured serum bilirubin levels (Fig. 2A
); similarly, acute
liver damage is accompanied by the release into the blood of hepatocyte
cytosolic enzymes the liver transaminases, including ALT (Fig. 2B
), and of hypoglycemia (Fig. 2C
). Most blood
coagulation factors are produced in the liver, and severe acute liver
disease is known to lead to a reduction in the production of
coagulation factors (1)
. Since both the coagulation
factors V and VII have a short half-lives (t1/2 < 8 h), we
chose to follow the synthetic function of the liver by measuring the
levels of coagulation factors V (Fig. 2D
) and VII (Fig. 2E
). When D-gal-treated rats were subsequently treated with
either glucose or IL-6, bilirubin levels remained high (>60 mmol/l)
(Fig. 2A
), as did ALT levels (>6000 IU/l) (Fig. 2B
) and, to some extent, hypoglycemia (
65 mg%) (Fig. 2C
), indicating that there was no relief of the severe
hepatotoxicity. Furthermore, the production of both coagulation factors
V (Fig. 2D
) and VII (Fig. 2E
) decreased to below
10% of the normal activity level. Note that from day 5 most of these
liver functions returned to normal, suggesting that the effect of the
D-gal in this model is transient and is followed by spontaneous liver
regeneration. In contrast to the liver functions of the IL-6 and
glucose-treated groups, there was no indication of severe
hepatotoxicity in the hyper-IL-6-treated group (Fig. 2A
, B
, C
, D
, E
).
We observed similar results when we repeated this experiment three
times. Thus, it appears that hyper-IL-6 reversed the toxic effect of
D-gal. We hypothesized that the reversal of the toxic effect of D-gal
by hyper-IL-6 might take place by the induction of liver regeneration.
The effect of hyper-IL-6 on survival of rats with FHF
Having ascertained that HIL-6 reversed the toxic effects of D-gal
on liver function (Fig. 2
), we wanted to test whether HIL-6
could improve survival after the administration of high lethal doses of
D-gal to induce FHF. We injected D-gal at a dosage of 1.4 mg/kg body
weight to male Fischer rats and after 7 h administered a single
injection of hyper-IL-6, IL-6, or glucose (see legend to Fig. 3
). All
the animals treated with IL-6 or glucose died after 2.5 days (Fig. 3B
). Two of the seven animals treated with hyper-IL-6
survived for more than 14 days. We understood these results to suggest
that hyper-IL-6 could induce liver regeneration even in a highly
stringent FHF rat model.
Hyper-IL-6 but not IL-6 improves survival of female Fischer rats
with FHF
To confirm that the results of the experiments with male Fisher
rats were not sex biased, we repeated the experiments using female
Fischer rats. Seven hours after i.p. D-gal injection at a dosage of 300
mg/kg body weight to induce FHF, these female rats were treated with
hyper-IL-6, IL-6, or glucose (dose administered as in previous
experiments). After 24 h, all the rats in all the treatment groups
were still alive (Fig. 4
). However, by day 3 only 1 of the 10 control animals survived (glucose-
and IL-6-treated groups). In contrast, 4 of the 5 animals treated with
hyper-IL-6 were still alive. These results supported our previous
results showing that hyper-IL-6 can prevent rat death caused by liver
function failure. We then wanted to explore our hypothesis that
hyper-IL-6 acts by stimulating hepatocyte proliferation.
|
The effect of hyper-IL-6 on hepatocyte proliferation
We examined the livers from female Fischer rats treated first with
D-gal and then with hyper-IL-6, IL-6, or glucose as described above
(Fig. 4)
; after 2 or 3 days, the rats were pulsed with BrdU for
immunohistological staining (Fig. 5
). We made more than 10 low-power field microscopic examinations. One
day after D-gal injection, 4050% of the hepatocytes in the
hyper-IL-6-treated group were BrdU positive (Fig. 5C
, F
). However, in the groups treated with IL-6 (Fig. 5B, E
) or glucose (Fig. 5A, D
) there were fewer
than 10% BrdU-positive cells. The results for each group were similar
on day 3 (data not shown). Thus, as we had predicted, the therapeutic
capacity of hyper-IL-6 is its potential to salvage from death animals
suffering from FHF.
The early effect of hyper-IL-6 in D-gal-induced FHF
In our initial experiments we observed the effects of hyper-IL-6
on D-gal-induced hepatotoxicity (Fig. 2)
during the first 40 h
after D-gal injection. We were interested in defining the earliest
point at which the physiological effects of treatment by hyper-IL-6
would commence. Severe hepatotoxicity was induced with D-gal, and
treatments with hyper-IL-6, IL-6, or glucose were administered
according to the protocol described above. In animals treated with
hyper-IL-6 5 h after hyper-IL-6 treatment (that is, 12 h
after D-gal administration), the activity of factor V started to
increase (Fig. 6A
). At the end of the first day, the activity levels of both coagulation
factors V and VII became normalized (Fig. 6A
, B
). In
contrast, at the end of the first day the activities of the coagulation
factors V and VII in both control groups remained very low. This
suggests that the therapeutic effect of hyper-IL-6 starts shortly after
administration, which could have significant ramifications clinical
applications.
|
We further evaluated the potential of HIL-6 to reverse the
hepatotoxicity effect of D-gal in the short term. We observed increased
ALT levels (Fig. 6C
) in all the groups to which D-gal was
administered. However, this effect was reversed 24 h later if the
animals were treated with HIL-6. On day 3, histological examination of
livers of the glucose- and IL-6-treated animals (data not shown)
revealed severe liver pathology similar to the changes shown in
Fig. 7A
, B
.
|
As an additional control, we compared animals treated with D-gal to
those treated with saline, then 7 h later with glucose, IL-6, or
HIL-6. In these saline-treated rats, we observed no increase in the
levels of ALT (Fig. 6C
) or bilirubin (data not shown) nor
did our histological examination reveal signs of liver dysfunction
(data not shown). Two days after treatment with saline instead of
D-gal, the level of BrdU incorporation in all three groups was the same
and observed in less than 1% of the cells (data not shown).
Hyper-IL-6 rescues rats suffering from TAA-induced FHF
Rats in which FHF was induced by TAA were divided into three
treatment groups of 10 each. Thirty hours after the induction of FHF,
each group was treated i.p. with 5 ml 10% glucose, IL-6 (80
µg/animal), or hyper-IL-6 (8 µg/animal). Note that the amount of
hyper-IL-6 administered per animal was 10-fold less than the amount of
IL-6 administered per animal. On day 4 after induction of FHF, 1 of the
10 animals treated with glucose survived, 2 of the 10 treated with IL-6
survived, and 4 of the 10 treated with HIL-6 survived (data not shown).
These results further support the hypothesis that treatment by
hyper-IL-6 is preferable to treatment by IL-6.
The effect of HIL-6 on D-gal-induced apoptosis
D-gal is known to induce significant apoptosis, although the exact
mechanism for this effect is not known (22)
. As expected,
we found that 2 days after rats were treated with HIL-6 after the
induction of severe hepatotoxicity, there was a significantly lower
level of apoptotic cells than when the animals were treated with IL-6
(Fig. 8
) or glucose (data not shown).
|
| DISCUSSION |
|---|
|
|
|---|
. Together they form a complex that
induces the homodimerization of two gp130 signal transduction molecules
(20
(sIL-6R) causes
IL-6R
-/gp130+ cells to be responsive to IL-6 (25)
B, suggesting a role for GP130 signaling in liver regeneration
(9
Here we report on the in vivo ability of the hyper-IL-6
cytokine to reverse both sublethal and lethal hepatotoxic damage
induced by D-gal in rats. In our experiments, neither glucose nor IL-6
could replace hyper-IL-6. When the D-gal-induced liver damage was
sublethal, liver cells started to proliferate shortly after treatment
with hyper-IL-6 and the toxic injury was reversed in less than 24 h (Fig. 6)
. The recovery in rats treated with IL-6 or glucose occurred
simultaneously 24 to 48 h later, suggesting that this delayed
event was spontaneous, possibly related to endogenous generation of
IL-6 and sIL-6R or a panel of cytokines and growth factors. Note,
however, that 10-fold more IL-6 (80 µg) was used than hyper-IL-6 (8
µg). Moreover, we found only minor pathological alterations the liver
histology of the hyper-IL-6-treated group, whereas the livers of
animals in the two control groups revealed severe liver damage. These
changes persisted in the livers of the IL-6- and glucose-treated rats
and were detected on day 3 (Fig. 7A
, B
). In contrast, at the
same time we found only minor pathological alterations in the
hyper-IL-6-treated animals (Fig. 7C
, D
), suggesting that the
therapeutic effect of hyper-IL-6 is achieved through enhanced
hepatocyte proliferation. This early recovery of the hyper-IL-6-treated
group we observed histologically coincided with a complete recovery of
hepatocyte synthetic functions, as apparent from the serum activity
levels of coagulation factors V and VII and normalization of the levels
of serum bilirubin, glucose, and ALT (Figs. 2
and 6)
. Furthermore, in
all three of our stringent models, treatment with hyper-IL-6
significantly improved the survival rate of rats with D-gal-induced
FHF.
Although at this time we can only speculate on the mechanism by which
hyper-IL-6 reverses liver toxicity, we propose a hypothetical algorithm
for this effect: D-gal might induce apoptosis by sensitizing the liver
cells to TNF-
(29
, 30)
. The designer cytokine
hyper-IL-6 is a potent activator of the signal transducer and an
activator of transcription STAT3 (31)
. Both in
vitro and in vivo, hyper-IL-6 is more than 10-fold more
effective than IL-6 in stimulating STAT3-dependent gene transcription
in liver cells. STAT3 is known to be involved in liver regeneration and
is also anti-apoptotic (32
, 33)
. Together, these facts
strongly suggest that the activating the gp130 signal transduction
pathway may induce enhanced cell proliferation. In fact, we observed
that HIL-6 did have a strong anti-apoptotic effect (Fig. 8)
. It is
possible that this effect might be induced by HIL-6 gp130 signaling the
activation of STAT-3 (34)
. Thus, based on these ideas and
on the results that we have reported here, we propose that hyper-IL-6
has an anti-apoptotic effect that reverses the D-gal hepatotoxicity and
enhances liver regeneration. It is also known that patients with liver
damage have increased serum IL-6. IL-6 stimulates the production of
acute-phase proteins (e.g., haptoglobin), and these proteins have
recently been found to induce the shedding of IL-6R from macrophages
(35)
. The resulting increase in blood levels of sIL-6R
would thereby be expected to enhance the direct effect of IL-6 on liver
regeneration.
We have shown that the fusion protein hyper-IL-6 has a marked stimulating effect on liver regeneration in rat FHF and severe hepatotoxicity models. Our results point both to an important role for sIL-6R during liver regeneration and suggest that hyper-IL-6 has a significant therapeutic potential whereby it could be used to reverse the state of FHF or to enhance liver regeneration in various clinical conditions related to acute or chronic liver damage.
| ACKNOWLEDGMENTS |
|---|
Received for publication October 20, 1999.
Revision received April 13, 2000.
| REFERENCES |
|---|
|
|
|---|
levels and inhibiting adhesion of T lymphocytes to extracellular matrix. J. Hepatol. 29,60-67[Medline]
B in the liver: the good and bad news. Hepatology 27,1445-1446[Medline]
B in activated rat hepatic stellate cells by proteasome inhibitors and an I
B super-repressor. Hepatology 27,1285-1295[Medline]
This article has been cited by other articles:
![]() |
U. M. Apte, R. Mcree, and S. K. Ramaiah Hepatocyte Proliferation is the Possible Mechanism for the Transient Decrease in Liver Injury During Steatosis Stage of Alcoholic Liver Disease Toxicol Pathol, August 1, 2004; 32(5): 567 - 576. [Abstract] [PDF] |
||||
![]() |
R. Bang, G. Sass, A. K. Kiemer, A. M. Vollmar, W. L. Neuhuber, and G. Tiegs Neurokinin-1 Receptor Antagonists CP-96,345 and L-733,060 Protect Mice from Cytokine-Mediated Liver Injury J. Pharmacol. Exp. Ther., April 1, 2003; 305(1): 31 - 39. [Abstract] [Full Text] |
||||
![]() |
U. M. Apte, P. B. Limaye, D. Desaiah, T. J. Bucci, A. Warbritton, and H. M. Mehendale Mechanisms of Increased Liver Tissue Repair and Survival in Diet-Restricted Rats Treated with Equitoxic Doses of Thioacetamide Toxicol. Sci., April 1, 2003; 72(2): 272 - 282. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Wuestefeld, C. Klein, K. L. Streetz, U. Betz, J. Lauber, J. Buer, M. P. Manns, W. Muller, and C. Trautwein Interleukin-6/Glycoprotein 130-dependent Pathways Are Protective during Liver Regeneration J. Biol. Chem., March 21, 2003; 278(13): 11281 - 11288. [Abstract] [Full Text] [PDF] |
||||
![]() |
U. M. Apte, P. B. Limaye, S. K. Ramaiah, V. S. Vaidya, T. J. Bucci, A. Warbritton, and H. M. Mehendale Upregulated Promitogenic Signaling via Cytokines and Growth Factors: Potential Mechanism of Robust Liver Tissue Repair in Calorie-Restricted Rats upon Toxic Challenge Toxicol. Sci., October 1, 2002; 69(2): 448 - 459. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |